Implementing a Multiagent-Based Self-Managing Structural Health Monitoring System on a Wind Turbine

Deteriorations, increasing costs and burdens of maintenance work as well as recent collapses of civil engineering structures underscore the need for robust, cost-effective structural health monitoring (SHM) strategies. This paper describes the design and prototype implementation of a multi-agent software paradigm for a self-managing structural health monitoring system. As will be highlighted in the paper, one of the key demonstrations of the prototype system is the timely detection of system malfunctions such as breakdowns of sensors, temporarily unavailable data acquisition units or server systems. Without such selfdiagnostic and management capabilities as in current conventional monitoring systems, valuable monitoring data – necessary for an accurate and reliable safety assessment – can be lost as system malfunctions are usually not detected at an early stage. The performance of the prototype is validated through a long-term system deployment test on a 500 kW wind turbine in Germany.